Burningbird

Hey Shelley (spelled it right this time!),

We've interacted before on your blog, and you probably found my personality about the same, but you didn't point the "ego" finger. Please don't armchair-analyze me now. ES4 and Mozilla's work on open standards is not about me or my ego.

If I were really that egotistical, I would be defending JS as it is, even when its flaws can't be worked around by clever closure tricks (which come at a cost). And if I had more sense than stubbornness (a fair cop), I wouldn't be at Mozilla trying to roll this "evolve JS" stone up a hill. But I really do believe JS should evolve for its users' sakes, not for mine.

There are obviously many egos, or I mean people, involved in TG1. That's life, but what is more important than any one ego or agenda is whether the web standards, definitely including JavaScript 1 (ES3), will be competitive in two or three years.

If you think Flash is to be avoided, and Silverlight was an unnecessary invention (to try to kill Flash? Let's just say to help people make better presentations and multimedia content, which should be possible with web standards), what do you think is the likely consequence of keeping JavaScript in its growth-restricted current form?

It's telling that both Adobe and Microsoft successfully promulgate stronger programming languages against JavaScript, when marketing Flash and Silverlight against the Web. This is not just snake oil, and chess demos aside (C# is much faster than IE JScript!), there are good reasons to want to evolve JavaScript for programming in the large, stronger mutation controls (important for security), optional types, and the other facilities in ES4.

Simplicity is not underrated in my book, but there's a wise line from Matz in "Beautiful Code" that I've been citing lately:

“Simplicity is one of the most misunderstood concepts in programming. People who design languages frequently want to keep those languages simple and clean. While the sentiment is noble, doing this can make programs written in that language more complex.”

Guy Steele has a great talk (from OOPSLA '98 IIRC) on this point, called "Growing a Language". Recommended.

/be

That money quote from the comments is from Douglas Crockford, not Crawford. As somebody whose last name is misspelled several times a day, things like that jump out at me ;-)

That being said, I echo your sentiments 100%. I'm not sure whose side I am on this time, either. I have long been a supporter of Mozilla (user since Milestone 19, if anyone remembers those long ago days), not least because of their commitment to standards.

So I was a little bit taken aback to read on Asa Dotzler's blog the other day that things proposed in the WHATWG's "HTML5" were "standards" and thus should start being implemented right away, since HTML5 hasn't gone through any formal standards process — not W3C, not IETF, not ECMA, not nuthin'. (The W3C is using it as the basis for the next version of HTML, but that process is just beginning - who knows how the final W3C spec will differ from the WHATWG's.)

I left a comment on his post to that effect and got a dismissive "the W3C isn't relevant anymore, WHATWG is" as a response. Which raises the question of why they're bothering to move HTML5 through the W3C in the first place. I know the W3C process is screwed up six ways to Sunday, but it at least has Microsoft participating, which is more than you can say for the WHATWG. Pushing WHATWG as a "standards committee" when its membership is made up of low-market-share browser vendors and nobody else strikes me as a recipe for more problems down the road, not less.

Fun times! :-)

Hi Jason,

My 2 cents, not speaking for Asa or saying he was right. Standards bodies are in trouble. The OOXML mess is just one recent example.

The W3C rechartered the HTML WG under moral pressure from web developers and competitive pressure from the WHAT-WG. We WHAT-WG founders always hoped to land our proposed standards in the W3C. But if the W3C HTML WG does not function, and there are warning signs, then what? De-facto standards wars give the dominant player too much power and make an interop mess for developers.

I don't have an easy answer, and I share your concern. But from what I can see and hear, all is not well in the W3C HTML WG.

Whatever happens, web developers and at least the minority-share browser vendors need something like the WHAT-WG to make forward progress through incremental improvements. Mozilla is going to keep participating in both the WHAT-WG and the W3C.

/be

Brendan,

thanks for the response — I appreciate hearing a voice from the trenches, as it were.

I think the WHATWG vendors absolutely did us all a service by getting the HTML process moving again. It was very disappointing to see the W3C letting it languish for so long, and very heartening to see the WHATWG's pressure get it moving again — especially with Microsoft on board. And the HTML5 spec itself was pretty solid and exciting too.

That being said, I think you're right that the HTML-WG process has gone off the rails, but I don't know if that's 100% the W3C's fault. When I heard that interested parties could volunteer to join the HTML-WG as "invited experts", I actually signed up — I had no delusions that I would have a big impact on the final product, since I don't get paid to work on things like that, but as someone who believes strongly in the value of an open Web I wanted to do whatever I could to help out.

Once I was on the mailing list, though, it quickly became nearly impossible to find any way to help; the people from the WHATWG vendors routinely (and aggressively) shot down ideas from anyone who wasn't one of them. There were many, many threads that started with somebody saying "What about X" and ended with a WHATWG person saying "we resolved that already, you don't know what you're talking about, don't ask questions."

The WHATWG participants made it very clear from the beginning that they had their spec already, and if anyone pushed on it too hard, they'd just walk out and implement it in their browsers without us. There was a very strong "butt out" vibe coming off them.

I'm not saying that all the ideas were good ones - obviously they weren't, especially since the group was completely open - but I had hoped that the process would be something more interactive and collaborative than it turned out the WHATWG representatives wanted. It was disappointing.

I eventually gave up and stopped trying to follow the threads, which embarrasses me, since I do care about this stuff and I want to see the final product be as good as it can be. But I was so worn out and turned off by the process that I decided to leave the bickering to the professionals. I imagine there were other well-meaning people who wanted to contribute too who were similarly driven off. It's a shame, because I think the final product would have been better if it the major participants had engaged all that energy, rather than dampening it.

I've really not followed the HTML5 stuff closely, but my impressions match Jason's. I've been disappointed about two issues: 1) the backsliding on XML, and b) the rather hostile reaction against incorporating RDFa (since microformats are NOT an adequate long-term solution to the problem of embedding metadata).

First, Jason, thanks for correction on name. My first name is usually misspelled, so I understand.

Brendan, everyone involved in the creation process has their ego involved, no matter how much they try to stay dispassionate. That's why I sometimes think group efforts should be led by the person who has the least involvement or dependency on the results.

Your weblog post can only be interpreted as a shot across the Microsoft bow. Perhaps this is business as usual within the group, and no one takes it amiss. I know for myself, I would have a difficult time continuing to work in a group that generates such sentiments. I can understand frustration, but as lead of the effort, you don't have the luxury to just let loose.

As for language, like I said, I'm not an expert in language design. However, for all the crap that the current version of JS gets, how widely used is it? If making it better increases the complexity of using the language, I have to ask: is the improvement for everyone? Or just language purists? I agree that security is number one. But the security issues I've seen with JS have been primarily involved with how the language is used, not as a nature of the language itself. Better education about security problems seems to be the more appropriate approach.

However, like I said, though, I'm not a language design expert.

Regardless of growing languages, if Microsoft pulls out, the issue is moot. Browser wars. Discrepancies. Conditional tests. Cross-browser libraries. I didn't have fun with these a decade ago. I can't see how they would suddenly be fun now.

It's frustrating because Microsoft really is being a butt hole — I can believe it. At the same time, no one browser has more market share than IE. We can't just blow off Microsoft. To be honest, I think we'd be better to do nothing, though that sounds harsh.

We can't afford another go around of widely different JS implementations in the major browsers.

As for HTML5, I also have gotten the impression that things are not going well. Not to mention confusing signals about what is or is not the state of HTML5. I get the impression outside commentary is welcome one moment, not the next.

I also agree with you, Bruce: I think it was a huge mistake to back off on XHTML. Now, everyone is confused about the direction to take. Again, rather than a new language, education, better tool support…and on and on.

As for HTML5 and microformats — you might as well call HTML5, "Google's HTML" because that's what it looks like to me. In my opinion, it seems that many of the microformats are for search engine optimization. At least RDFa would have enabled broader metadata support.

Don't even get me started on Canvas.

I respect the hard work folks like you, Brendan, and Hickson with HTML5 (as well as other standard/spec work), do. At the same time, all of what happens in these efforts doesn't impact on just the browser makers, and search engine companies.

Re: "I also agree with you, Bruce: I think it was a huge mistake to back off on XHTML. Now, everyone is confused about the direction to take."

Well, I'm frankly torn. If I could, I'd use XHTML5 + RDFa, since the former has some of the richer structural support of XHTML 2, and the latter has a real metadata model.

But given the politics of all this, I have a suspicion that it won't be a reasonable option (despite the fact that it is technically trivial to do; just allow foreign attributes on any XHTML 5 element).

I will never use HTML 5, nor will I waste any more of my time on microformats development. But yes, the signals are confusing, and counterproductive. To me the discussion of embedding SVG in HTML, and of XML 5, are just insane.

Shelley, I was wondering what your reference to Al Billings' post was referring to; at no time did I say "we're picking up our marbles and going home"?

Chris, I picked this up from Al's statement:

"I gather from Chris’ post that, while they aren’t going to out and out say so, it is very unlikely that IE will support ECMAScript 4 in IE8 (or perhaps even beyond that)."

You didn't disagree with this in the comments. Sans communication of direction MS is taking, one has to assume we're heading into another round of scripting incompatibility. Not to mention 2D graphics incompatibility.

You may not have said the words, but you sure dropped this impression across several places.

Shelley, my blog post is a return shot. The first two links point to prior shots from Chris. Fair's fair, and I'm under no obligation to be meek.

But thank goodness we opened up all the TG1 materials via that wiki! Otherwise it would be he-said/he-said. Truly open proposed-standard development and transparency are good things.

On the charge that I'm too feisty to be the Convenor of TG1, I'll let others put me on trial. If it's time to find a replacement from a more neutral corner, so be it — but that will not lead to "squaring the circle" or "finding a middle way", from everything we have heard.

TG1 should certainly try to cut ES4 down with some care, and I hope we do shrink it. But halving it through some mediation process, simply because the minority wants none of it while the majority wants all of it, will please neither group and not result in good language design. I hope you can see this.

The reason I supported the "ES3.1" effort as an activity within TG1 was precisely to avoid an incompatible fork. Any "3.1" (never mind Ecma version numbering limitations) should be forward as well as backward compatible.

But it seems the 3.1 work from March and April is a dead letter, and we're hearing now about a new language being developed in secret by Doug Crockford and Microsoft.

/be

Brendan, I'll have to ask you, in your own turn, not to spread lies. There is no "new language being developed in secret by Doug Crockford and Microsoft" as far as I know, and I would be in a position to know. I don't know where you got that impression; it is false.

Additionally, characterizing my opinion, at least, if not Microsoft's (I won't presume to speak for Douglas) as "wanting none of it" is also patently untrue.

Shelley, I pretty much agree that the business case for me and you to use microformats is search, and that's really about it. Microformats were developed by search engine guys and originally supported by a search engine company. For search engine companies, microformats make it easier to scrape data.

Hmmm a lot of discussions about HTML 5. I'm very careful about people expressing themselves through weblog posts, emails, comments, because we are assuming a lot about the other person intents in the writings.

But I'm surprised to see the repeated motto from parts of the Web community to say that "standards bodies are in troubles". I'm working for W3C apart of that I'm not foolish about what is a standard body. It is not a church or anything else like this. It is a community of people which at a point decide to work together.

For the long story of HTML, the move to XHTML is not something new, the first mentions and discussions about the evolution of the Web are in the minutes of a workshop in 1998 before the famous workshop on Web applications. The funny thing reading the minutes is that you can read exactly the same kind of comments we can read on the html wg mailing list now.

For participation and process, it is not easy to work collectively in a group, when each time a proposal is made, a part of the groups says if it's not going the way we want we will work by ourselves. Or from Ian Hickson himself: "I'm not interested to work on a document which is not for Web browsers." Sic. Working together… is not about working on what please oneself only.

But as I said all of that is online communications, and I'm willing to meet some of these people at the Technical Plenary next week. Though some have difficulties to interact in a real™ human relationship. That will be interesting to see what is happening.

The thing which is worrying that people prefer to make big statements about Process, more than trying to work together on a solution. I have never met yet Brendan Eich, I hope I will one day just for the sake of my own enlightment.

Chris: Doug Crockford says Microsoft is involved in a new language proposal being developed jointly with him, and so far in private (not available to TG1 or the public). Quoting Doug: "I am pursuing with Microsoft a counter proposal for a simpler, reliable remedy to real problems."

I leave it to you to talk to whomever at Microsoft Doug is working with.

If you want only some of ES4, which parts? The last we heard in TG1 was from Pratap Lakshman in September: "Our spec is an edit of the ES3 source document itself". From those meeting minutes, it's clear there is some kind of proposal based on ES3, still not available to TG1. It is not ES4, and what it might lift from ES4 is unknown. If it's the old "ES3.1" proposal from the wiki, why has it not been maintained there?

If you feel the urge to counter my blogged objection to your repeating lies (that I shouted down dissent or ignored objections in TG1) by accusing me of lying about something, you might want to do it on a subject that is not documented on the public wiki by TG1 meeting notes editable and vetted by Microsoft's reps.

I've cited my sources. I'm not lying about Doug and Microsoft telling TG1 members that they have a new proposal coming. When will we see this new, so-far secret spec?

/be

I think Brendan Eich is right. I've had the same experiences he has with regard to Microsoft and how they like to do standards development. The fact that they now develop an alternative to ES4 within the physical and virtual walls of Microsoft with absolutely no communication to or from the outside world (except Mr. Crockford) just underscores this even more. Microsoft likes to develop a specification in-house and then deliver it done baked to a standards organization. They do not like to work together in the open with the community. This is exactly how they are doing "Open Source" development too, by the way. They are good at putting stuff out there, but they are terrible at taking stuff back in.

What I don't agree with Eich about is how he express himself, which is what I think you, Shelley, too, had a reaction from. Using the word "lie" like Eich does (and repeats in comments here) just isn't very constructive. It doesn't really matter how right he is about it, it's just not something you paint a wall with. Let people read between the lines instead; that way you can get the message across and not have to retort to pissing at the opponent.

What I think is important to keep in mind when discussing JavaScript, ECMAScript, et al, is that the language doesn't only live in a web browser. I agree that the web browser is its primary execution environment (today), but it has many other environments to be executed in that also gain a lot from being moved forward in the direction of ES4.

I actually think ES4 is a superb specification. It's backwards compatible with ES3, so all existing ES3 code will still work in ES4. What we need is just a good migration path to ES4 so ES4 code won't make old browsers go bananas for the next couple of years. When we have that figured out, I don't see a problem with ES4 at all. Except of the minor disturbance it is that Microsoft probably won't implement it, of course.

Brendan: "proposal" != "language proposal". Proposal for what priorities should be, absolutely; proposals for specific areas in (or not in) ES3 or ES4 that should be addressed, absolutely; but not a "here's you're language spec" proposal.

Perhaps if suggestions were not met with this kind of hostility, Douglas, Pratap et al would feel comfortable having these discussions directly in the TG1 meetings, since as you imply, secrecy serves no one.

Chris, you wrote on the IE blog that "a revolution in ECMAScript would be best done with an entirely new language". Are you saying that language doesn't exist, and the ECMAScript committee must stop work on ES4 none the less, because you don't want to implement it? That doesn't wash.

I think it's pretty bogus to declare that TG1 must switch its focus. If you want 3.1-type stuff in part of ES4, the committee has always been open to that. It is the original agreement with Pratap and Doug (it's in the notes, do not argue). No one is forcing you to implement all of the spec (but we would like it if you did).

'On the charge that I'm too feisty to be the Convenor of TG1, I'll let others put me on trial. If it's time to find a replacement from a more neutral corner, so be it — but that will not lead to "squaring the circle" or "finding a middle way", from everything we have heard.'

Brendan, frankly I don't know if you're too feisty or not. I'm just commenting on the communication I'm seeing. As for not making a difference in this regard, fair enough.

As long as everyone here remembers that the whole point of having a standards effort for ECMAScript was so that we don't have to spend the next decade working through browser differences.

If it takes dueling at twenty paces to accomplish this goal, I'll count off the paces, drop the handkerchief, help mop up blood after–and refrain from making comments about "Liar! Liar! Pants on Fire!"

Shelley, I think you're overplaying it. Chris posted factually incorrect information on my weblog, and Brendan corrected it. He did not call Chris a liar. However, the information did come from someone (don't know who), and that person was lying.

Now Chris is claiming that "suggestions" are "met with hostility". The only suggestion I see is "stop what you're doing and let the minority dictate the agenda of the committee". Sorry, no. :)

Brendan told Chris to stop spreading lies, Robert. I may call it pink and you call it rose, but it's still a shade of red.

My take on all of this, trying to read through everything that's been pointed out, is both sides of this issue are more closely aligned than not.

No one wants to break backwards compatibility with previous versions of JavaScript. The current ECMAScript effort is trying to add additional programming constructs. Microsoft doesn't want to implement these additional constructs in JS, because Microsoft wants to go another language route. Going another language route seems to be something that Mozilla, and other organizations also want to do.

I imagine that MS doesn't want to be accused of not 'supporting standards' if it doesn't implement these new extensions. That's understandable. Everyone beats up on Microsoft, and sometimes, it's not deserved.

(Well, as regards to 2D, it is deserved. Why can't the same people who gave us Silverlight give us SVG and Canvas? I digress…)

What's needed is a road map showing how there's a fork in the path in client-side language support up ahead, and one path goes a JS specific route, another a common language runtime (or Python-based or, hopefully not, Ruby-based) route, but all browsers promise to support at least one of the paths going forward: either the JS route at a minimum or a truly common language runtime approach, or both. Regardless, both provide a common level of support for all of ES3. And there's no harm if one browser maker supports one over the other. No, we won't beat up on MS if it chooses the different language approach.

As long as we're agreed that there's two paths — not one for each vendor, each browser, and let's throw in one for Google, just because.

All parties involved should keep one thing in mind: the purity of the language doesn't matter. What matters is the usability. Language purists may not like JavaScript, but it certainly has shown itself to be usable. Client-side development has been a very open environment.

Don't muck it up.

One last PS:

I respect all people involved in the discussion here and elsewhere. I don't think people are lying. I think people are hampered by corporate hushups that may not hamper others. I don't think people are being mean–I think they're frustrated.

But, I'm not part of the working group.

When I wrote I was disappointed, it was because this effort has broken down into Chris and Brendan running around in comments on various posts, telling each other to either stop lying, or to stop being so mean.

Going another language route seems to be something that Mozilla, and other organizations also want to do.

Not sure what you mean. ES4 extends ES3, it isn't a new language. In fact, you call them extensions in the paragraph below. Not everyone likes every extension, but you're not forced to use any of them.

I imagine that MS doesn't want to be accused of not 'supporting standards'.

Huh? The ES4 work is an agreement between Adobe, Opera, Mozilla, and others in a standards organization. We call those standards. If Microsoft doesn't implement ES4, that will be a standard they don't support. I don't see the appeal of being wishy-washy about it.

Let's stick to the facts that can be checked using open sources:

* Chris said Ecma TG1 including its Convenor (that's me) was ignoring and shouting down dissent. These are serious charges, they need evidence.

* I posted links to the ecmascript.org wiki's proposals, discussion, and meetings sections showing tons of free exchange between all TG1 members, dissent piled on top of dissent without shouting, and no ignoring except on the part of the dissenters, who stopped responding to open comments, or doing any other work in the wiki, in April.

Shelley, you may think I'm the worst offender for calling a falsehood knowingly told a lie. I think that your umbrage at me takes Chris less seriously than he deserves, and lets him off the hook. What he said is a serious charge against both Ecma TC39-TG1 and me personally. His charges require evidence, and since I've given plenty of contrary evidence, they ought to be retracted.

Putting manners over morals is upside-down (what's the epigram? "Manners are morals writ small"). In TG1 meetings and hosted dinners no one has shouted, no hostilities have erupted (this is not true in all standards bodies — IEEE 754r, for example, had shouting and table-pounding, and ended up using Robert's Rules!). The meeting notes and the dinner bills (at least mine) prove this.

So Chris's whining here about how I'm hostile to him, as if that proves I was hostile to anyone in TG1, is a lame attempt to project an incited reaction backward in time. If it convinces you, then I do not know what to do. As I wrote, I'm not the meek type. But so what? You ought to be more concerned about the substance instead of the surface.

/be

Brendan, I don't think you're an 'offender' at all. I also don't see Chris' expressing his opinion to be 'serious offenses'.

I think there's too much emotional context to this discussion and it might be good if all parties step back and take several long, deep breaths. This isn't manners. It's common sense.

I am sorry now that I wrote what I did about your use of 'lie', as it seems to have really hurt and offended you. I still believe in what I wrote, but the timing was poor on my part. I know better than to write what I did while emotions were still running so high. I apologize to both you and Chris for that. We on the outside have added much fuel to this fire, and it's time for us to keep our opinions to ourselves and let you both work this out, without input from the gallery.

Good luck with your effort Brendan and Chris, both. Since I'm keenly interesting in the future of JavaScript, as I know both of you are, I hope all of the issues can be resolved.

From the perspective of a web developer that spends a lot of time with object oriented server side code as well as a lot of time in current versions of Flash, I would much rather see ES4 as an upgrade to the current spec rather than as a separate language.

@Mr. Wilson I'm sorry to say, that, Silverlight not withstanding, I simply don't trust Microsoft to work within the constraints of a standard and to not come up with something that will be restricted by platform.

Your complaints about backward compatibility simply don't ring true to someone who has successfully made the transition from Actionscript 2 to Actionscript 3 and from PHP 3/4 to PHP 5. The very engine running AS 3 will be the engine running ES3/ES4 in Firefox. Most of the people posting here have more knowledge than I about the internals, but I am one of those that will be using these tools. I also have often had to occasion to answer questions on forums for new Javascript coders.

I can tell you, without a doubt that what would be far more confusing than a major upgrade, would be two separate run times for Javascript. I can teach someone OOP, but what you propose is going to be confusing and unworkable.

The issue of backward compatibility should be settled in my estimation.

Developers and end users are already starting to get concerned about bloat in browsers, and that is definitely an issue with mobile platforms. If we have another language, then fine make it a known quantity like Python or Ruby, and make it totally optional. I imagine Intranet developers and hobbyists may get use out of it. Creating the impression with developers that there is a basic and pro version of the main scripting language, that would need to be supported is absurd. This is my impression of what you propose, and I think most web developers would be opposed to it.

"At least RDFa would have enabled broader metadata support."

RDFa is technically inadequate, both structurally (qnames in attributes) and semantically ('about' referents are hazy).

> So Chris's whining here about how I'm hostile to him, as if that proves I was hostile to anyone in TG1, is a lame attempt to project an incited reaction backward in time. If it convinces you, then I do not know what to do. As I wrote, I'm not the meek type. But so what?

Can't you see that the attitude exemplified by this comment is precisely what Shelly is talking about? This is not how a leader behaves, this is you being hostile and unproductive. If you have a problem with what Chris said, then by all means say that you think that he was unjustified and ask him to substantiate his claims. But *don't* call him a whiner and a liar. That is shouting him down, precisely what you claim you haven't done!

Quite frankly, I have a great deal of resentment built up towards Microsoft, but if this is the attitude that they have to deal with to work with you, then I'm not surprised that they decided not to, and each further comment you make on the matter convinces me more that Chris was spot on with his characterisation of you.

Jim: I didn't call Chris a liar. Please read more carefully before you inveigh against me. Which is worse, calling a falsehood that was knowingly told to Chris a lie, or your misstatement of what I wrote?

As for whining, sorry — I didn't write that in a hostile mood so much as a mocking one. Please do try to tell the difference :-P. I'd much rather people speak precisely and say what they mean, than muffle and waffle about things. Spin campaigns thrive on imprecision.

/be